I probably bit more than I could chew, translation, building the Geo-Database for Sampa was probably the hardest feature to date on Sampa. There are many aspects that make it harder, but the number is probably the lack of consistency on terms of geographical names and cataloging them.
I want a Geo-database to do two things:
Given a geographic name (city, state, square, river, etc.) return an unique identifier for that location and its coordinates.
Given a coordinate, find the most likely geographic entity that this point belongs to.
For #1 I could have used your traditional geo-coder (from Yahoo, Google, ServiceObjects, etc.), but the problem is that they don't return a unique identifier. Using the city name is not a unique identifier because many cities have the same name, using a combination of the neighborhood + city + state + country would be unique, but it would be too long to store on each picture (for Geo-Tagging).
The problem with implementing #1 is the amount of duplicate names. For example, there are more than 40 places in the world named "São Paulo". In 99.999% of the cases, when somebody types São Paulo they are probably refering to my birth city and not the other tiny places around the world. Same thing for places like Paris, Tokyo, London. All of those names identify many places, but they are likely to be refering to "Paris, France", "Tokyo, Japan" and "London, England".
So, the only way to return the correct value in the correct order is to use some kind of Importance factor. In my view, and on the context of pictures/blogs, the things that make a place more likely to be what the user is looking for are:
The population of a location: Higher the population, more likely to be a match;
The size of location (as in how many square kilometers of area) -- Think of a national park or river (it has no population)
The number of tourists per year -- this is interesting because a lot of tourists hot spots don't have a lot of population, like Seychelles, Cannes, Venice, Angra dos Reis, etc.
To make the ranking more reliable, I manage to get the population of about 30,000 cities. But I'm still not happy, because my database has about 3 million locations.
How to get the size of locations? That is probably the hardest thing to get in terms of Geographic databases.
The number of tourists could be extrapolated based on the number of hotel's and/or the number of hotel rooms on that place. But where to get this data from?
I just got the database to work, the code in place and searching for "São Paulo" returned the big city in result #1, which is what I needed. I feel a bit relieved that the system works the way I designed it. I spent way too much time on this, while I should have re-dimensioned the feature to make it simpler. Two weeks are gone and just 1 feature gets implemented. Very frustrating.
I used to have this theory that every electronic should stop working after 5 years, so it forces you to upgrade. Computers, TVs, DVDs, MP3 players, etc.
However, for certain things it should be more like 3 years. Like Digital Cameras. The evolution of digital cameras is more impressive than the evolution of Microprocessors and Hard-disks.
My Canon Digital Rebel is showing its sign of aging. I bought it the same week that Canon released it at the end of 2003. It was the first worth having Digital SLR. When they released the 8MP version (I think it was last year), it wasn't enough of a difference for me to replace mine (w/ 6MP), but now with the 10MP the story starts to change.
The battery of my camera is almost dead. The noisy on low light it is bothering more nowadays (because they new cameras are better). And the flash memory that I use on this camera is so freakin' slow that is unbelievable.
But I'll need to wait at least until next year. Right now all my resources are going to Sampa.
Robert Sutton is publishing a book called "The No Asshole Rule" and Guy Kawasaki got an early edition of the book. Guy summarizes very good points from the book.
How to detect an asshole:
"...If you hear someone at a Starbucks order a 'decaf grande half-soy, half-low fat, iced vanilla, double-shot, gingerbread cappuccino, extra dry, light ice, with one Sweet-n’-Low and one NutraSweet,' you’re in the presence of an asshole. ..."
He also has many tips on how to avoid being an asshole and how to deal with assholes. Worth reading. Link.
Election time makes you think about a few things. When I first heard of the Electoral College in the US I was a bit surprise. How come something like the 2000 election could have happened? Where somebody wins the public vote, but loses the presidency.
My americans friends point out that this is to keep the system fair and avoid small states of not having any representation. I can't get over how stupid this is.
And then I realize another problem with elections, which is widespread on all countries of the world...
Why is representation by geography?
That is another thing that can be perceived as stupid.
Why people from Washington State gets to elect 9 representatives and people from Vermont gets to elect 1? Clearly, this is because the population of Vermont should be about 1/9th of the population of Washington. So far so good. But the laws being approved at the federal level are mostly about national or global issues, so why geography matters?
What if, instead of dividing the House of Representatives and the Senate by state, it would be divided by race, age, gender, religion, social class, etc.
Right now 13% of all Americans are Black, the Senate has 100 seats and the House has 435 Representatives. That means to be a true representation of the US, there should be about 13 black senators and 56 congressman on the house. Let's see...
In the House there are 40 black representatives (by my account) which is not bad, but on the Senate there is Barack Obama, and ... Nope, that is it. One black senator. Clearly, Whites are over-represented on both houses.
Now, we can use the same type of math for the other factors:
Age: this is the oldest congress in history with an average of 56 years-old for Representatives and 61 for Senators. The median american age: 35.3 years old.
Gender: Women should have a few more representatives, since 51% of the population is female, but according to Wikipedia, only 15% are congresswomen, while 85% of congressmen.
Religion, Social class, etc.
My suggestion is a new House of Representatives that is effectively a representation of the US population. Clearly you can't have a 7-years-old representing his age group, but besides some minor issues, you can get a pretty honest and representative system.
Religion: Catholic, Baptist, Methodist, Lutheran, Other Christian, Jewish, Others
Social class: A, B, C, D, E
Divide each voter into the same profile as above, so that each one can only vote for people that fit their profile. Each profile is considered 100% of the votes, which means if a black candidate gets 80% of the votes for black candidates, he ranks higher than a white candidate that gets 60% of the votes from white voters.
Stack rank all the candidates and you've built yourself a representative House of Representatives.
Now, there is some complicated logistic to make it work, but if it did, it would be awesome.
ImageKind is this cool website released recently by Kelly Smith (a friend of mine) and others.
If you read the "The Long Tail" you'll understand what I'm talking about. This is just a few of the reasons I think this will be a successful Web 2.0 company, in no particular order:
From day one, they understood the revenue model;
They are empowering all artists, professional or not, to create their own web-based art gallery and sell it directly to customers.
They totally enable the "long tail" effect (very low cost of carrying millions of items).
They understand the space (Adrian has a framing business).
I exchange many emails with Kelly while they were building ImageKind. They put a lot of effort into creating a great user experience.
They are creating a business eco-system (think eBay).
Of course you love your family, friends and co-workers (some of them, at least), but in the context of (your) business their opinion is very tainted. Let me explain better.
If you are embarking into the tough world of Startups, and you leave your safe and secure job to try to create something new, usually, the first thing that a reasonable person would do is to validate what he or she is going to do. And who is better to ask them the people around you? They are right there.
$&@#... that is you falling into a pit.
You asked the people that like you the most and are probably most afraid of offending you, or shattering your big dream. Why would they do that? Maybe you do have a couple of honest friends that will tell you the "truth", but that still is useless. Why? 'Cause they are not your customers!!!
Repeat 3 times: Ask the Customers! Ask the Customers! Ask the Customers!
When you hear your friend Mark telling you "this is the most brilliant idea I heard in a long time", or when your Entrepreneur Aunt Janice tells you "you've got what it takes to make this idea into a successful product" think twice. They might be experts, they might know the field, but are they your customer? And they can't say they will buy or use your product, you have to be sure they will.
People can say whatever they want, but once you ask for an upfront payment for the product to help with the R&D effort, if they don't put the money down, they are not your customers.
Now, on the context of keeping you motivated and support you during tough times, having the comforting words of Family and Friends supporting your idea is very important, but that is that. It won't help you get more customers if your product doesn't solve their needs.
About 10 days ago I remembered that I needed to go to Apple's website to find out if my battery was recalled. It had been a couple of months since the whole deal started and I have not received any letter.
So, on October 21 (Sunday) I went there, entered my laptop serial number and bingo. It was recalled. I filled in the info and receive an email confirmation.
Then, just 3 days later I received a letter from Apple telling me the battery had been recalled and I should go to the website. A-ha! I'm ahead of you -- been there, done that.
Yesterday, just 5 business days later, I received my new battery. I was pretty amazed since on the confirmation email they said it could take between 4 and 6 weeks. My luck was to go to the website before Apple mailed their hundreds of thousands of customers.
The only un-Apple thing in this whole process was the letter explaining how to return the old battery. I was two pages long! I'm sure they could have simplified and made it short.
I'll be going to the Entrepreneur University (EU) this Thursday. This event is organized by the Northwest Entrepreneur University (NWEN) every year and I attended the last 2 years. They used to be at the University of Washington, but this year it will be on the Sheraton Hotel in Seattle.
Of all the training, seminars, classes, breakfast meetings, and talks that I ever attended, the EU is by far the best. It is full of entrepreneurs, the presenters are fantistic people (most of them) from all realms of expertise to make an entrepreneur succeed. I end up meeting interesting people and you can always find a familiar face. Usually somebody from MSFT that is thinking about leaving. :)
Last year Glenn Kelman, the CEO of Redfin, just knocked it out of the park. His presentation was awesome (I tried to stalk him into looking into Sampa, but that didn't work).
The interesting (maybe obvious) thing is that you never know which presenters will be good or bad. Their status, years of experience, position, wealth and other factors are not indication of how good they are at giving a talk.
Well, I'll be there from 7AM until 8PM, so if you see a guy handing out free business cards that is me (or somebody else).
The US is by far the most entrepreneurial country in the world. I read somewhere its current status is mostly due to the healthy Venture Capital eco-system.
The core problems with the Venture Capital, IMHO, is that investment decisions are not made based on product merit. Sure, they won't invest in a very stupid idea, but they will invest in a mediocre idea if the "team" is a proven team, but that is not what happened to Yahoo, eBay, Google or YouTube. Completely unproven teams.
Just to be clear, VCs decision are very complex, but most will tell you that the experienced team trumps all other criterias, then comes other factors like market, opportunity, product, etc.
I just read (through Jeff Nolan on Venture Chronicles) that a VC from Boston (Charles River Ventures) is creating a new deal flow called QuickStart to make it easier for companies to get Convertible Debt (a.k.a. Bridge Loan) to get started. Fantastic. (Note to self: Send Sampa's business plan).
Here is the deal, nearly 100% of VCs will tell you they do Seed Investment -- it's a lie! Nearly 100% of VCs will tell you they can do Convertible Debt -- Oops, another lie.
Ask your favorite VC when was the last time they did a $250K Convertible Debt investment. Remember to exclude the one's that were done by EIRs (Entrepreneurs in Residence) because that is just you giving your son an allowance and telling people that you give money to the needy every week.
This is a great book from Chris Anderson, a Wired editor, describing the real inovations that they Internet is bringing to the market.
Having worked on MSN Search before, I have to say this is not big news. We always understood the power of the tail. At the time I worked on MSN, the "tail" was responsible for 2/3 of our traffic and revenue.
It is hard to define the "Tail" (since I'm not a writer), but it is easy to understand. In any business you have the "hits", products that sell way more than anything else. On MSN, the "hits" were the top search terms like "sex", "hotmail", "maps", "pictures", etc. The tails are the queries that occur just about a few times a day, like "haircut in redmond", or "how to fix canon battery flap".
In the context of Amazon, their hits are the bestseller, but the tail represents a huge portion of their sales. If you are in any business that has "hits", it certainly has a tail and you probably can tap into that to either grow your business, or to create a new business that is supplying a demand for that untapped market.
So I've got hundreds of log files from Sampa. It has been a problem to figure out how to use that data. We know what we want, but besides building a Log Analysis from scratch I wasn't able to find something good.
We tried AWStats, but we didn't like it. The reports are not nice and what I need is something that I can easily manipulate in Excel and/or import into a database to further analysis.
We want the basic stuff, aggregated for the day, week and month:
Fact: 99.99% of Internet users are nice people, just want to have fun when reading, posting or leaving a comment on the Internet. But then, there are 500,000,000 Internet users in the world, so, there are at least 50,000 idiots out there just waiting to screw with you.
I've been hit twice. The first time was because a friend of mine (Kelly Smith) got targeted by an idiot that thought he stolled his web site design (turns out that the designer used an 'open' template). I had nothing to do with it, but since Kelly had recently written about me and Sampa, people started bashing me as well. These usually occur in the form of threatening or very offensive emails and blog posts.
The second time I was hit was about a post where I wrote that I think using XHTML is a waste of time. Can you believe people would feel so annoyed by it that they would threaten me? But that is what happened. I've got about 50 comments/emails calling me all kinds of name and wishing me all kinds of bad stuff. (That is what you get to be on Digg's homepage).
Anyway, I was just reading Michael Arrington latest debacle (can I call it that, Michael?). He has been hit a few times. The only thing that I can think of is that this is making him stronger and with thicker skin. That is what happened to me.
The only thing I can suggest is for you to either get a thick skin and be prepared to receive absurds, offensive and threatening emails, or, just close your site, close your blog, take your pictures and videos out of the Internet and never share anything online.
Michael J. Fox is right -- No, I'm not talking about Stem Cell research (he is right about that too) -- I'm talking about the phrase "What you do in Missouri matters for millions of Americans". Nobody will question the factual validity of that statement (wait, some idiot will). A Senator from Missouri will be voting on the same laws that a senator from any other state, and most of those laws will affect the entire nation.
Now, I already said that "geography-based voting" is kind of wrong. So, when the US decides to invade or occupy a country, or when the US decides to apply importation taxes on another country's oranges so that local producers can better compete on price, or, when the US decides to build a fence on its border is a matter that affects the entire nation. Heck, it actually affects multiple nations. Sometimes the entire world is affected by US decisions. No question that US is the biggest (only) super power in the world today.
Wouldn't be fairer if the entire world would be able to vote for the US president and senators?
Now, to be fair, most of the decisions on Washington DC only affect US citizens (and foreigners living here), so, the entire world might have a weighted vote of, let's say, only 20%.
For example, if candidate A got 51% of the vote inside the US and the world voted 25% to candidate A, then the weighted total vote for candidate A is 45.8% (51%x80% + 25%x20%), which means that candidate B wins the election.
Now to be fair, this shouldn't be a change in the US election process only. Most countries have affects on their neighboors and trading partners, so the same type of logic some be used. Let's pick any country, like Belize in Central America. This is a small country that doesn't have a lot of international trade, so, less of the world vote should count, and probably only the direct partners and neighbors could have an influence.
Tell me, wouldn't that be much more fair?
PS: Note to self: Can we use PageRank to calculate the weight of each country vote? Better yet, let's do a PeopleRank where each people is influenced by each other and that amount of influence defines how much he can influence back on the original person's life.
Yesterday, it started a 7:30 AM in a raining day and it went through the entire day until 8:00 PM, non-stop. That is the Entrepreneur University organized by NWEN. It was an awesome event. Third time I attend and this is by far the best Entrepreneur event that I attend every year.
Here is a quick summary of my day and some of my thoughts.
8:00 AM: Morning Keynote by Jim Sinegal, founder and CEO of Costco: It was ok. He did a lot of advertisement for Costco and talked about their history. Not a lot of value but good to know anyway.
9:00 AM: Nick Hanauer (2nd Avenue Partners): Great, great talk on what he sees as transformational value created by a company. A few people probably left that talk and decided their idea was not good enough.
10:00 AM: David Weld (SeaTab software): It was supposed to be a talk about "Listen to Your Customer". It wasn't awful, but near awful. He brought a customer to talk about him. Boo-hoo.
11:00 AM: Jon Roberts (Ignition Partners): The talk was about scaling your business. And he starts with a bomb: "Don't scale to early", then he lists 9 things that you must achieve/be doing before you decide to scale. Very good.
12:00 PM: Lunch. I sat on the table with a guy from Marchex (Ryan Burt) with the theme "Web - Search Engine Marketing". Not interesting.
12:30 PM: VC Bistro: 5 minutes to talk to a VC 1-on-1 (about 10 of them). I picked Dan Rosen, Chairman of the Alliance of Angels. Great guy, great talk.
1:00 PM: Chase Franklin, founder and CEO of QPass: It was a good talk. He explained what the went through during the dot com bust (BTW, Chase, "dot com" should be spelled like that, and not "dot.com" because then it would be "dot-dot-com" )
2:00 PM: Therese Adlhoch Smith: Talk about Marketing on the Cheap. Similar presentation from another guy last year called ("Creating a brand on a shoe-string budget"). I hope someday these people understand that "cheap" and "shoe-string" is a couple of thousands of dollars, and not $50K or more!
3:00 PM: Denny Weston (Fluke Venture Partners) on "Writing an Executive Summary". It was a good talk. I knew most of the stuff there, but I'm sure a lot of people learned a few new things.
4:00 PM: Janis Machala (Paladin Partners) and Craig Sherman (Wilson Sonsini) on "Funding Early Stage Companies". This is probably that eleventh time that I hear either Janis or Craig (or some other Angel or Lawyer) talking about the same bullet points. Boring for me, but very valuable to other people.
5:00 PM: Pete Higgins (2nd Avenue Partners) on "Ten pitfalls for Early Stage companies": Best presentation of the day. I didn't know Pete was so good at giving talks. It was entertaining, informative and had a lot of value.
6:00 PM: Cocktail. Great for networking, which leads me to...
Geir Hansen: Manager at Silicon Valley Bank. He loans money to companies and VCs. Added an option to when you are about to close a round, but need immediate cash. I'll keep that in mind.
Hans Omli: I've met Hans at the TechCrunch Party Seattle. Mike Koss had introduced us by email before. He is just starting a new company. I'll talk more later.
(Microsoft Employee Number 12345): Can't tell his name because he still works at Microsoft. I've meet people like that every year. He is thinking about quiting and starting his own company. Good for you.
Andy Sack turned off comments on his blog. Actually, he didn't, but now people need to sign in with some stupid authentication mechanism. I never, ever sign in to leave a comment. Not because I want anonymity -- I always leave my real info -- but because I hate the hassle.
Usually, people do that because they been bombed with harsh, offensive or threatening comments, which I talk on this post.
UPDATE: My criticism is about not allowing comments on blogs. As simple as that. I'm not criticizing Judy's Book or Andy Sack per se. Apparently there are quite a few people that don't like Andy and Judy's Books.
Chris was my lawyer when I established Sampa. He and I lived just about a block from each other in Redmond (that was a coincidence). He was a great guy and had a terrible car accident in May and died a few weeks later. The sadest part is that he left two kids behind. Since my son was born in January I fell horrible when something like this happens.
Nick Hanauer from Second Avenue Partners gave a great talk on what he believes are the core traits of an entrepreneur, which are:
Ability to create transformational value;
Be a social outcast.
About #2 it is pretty much a fact: Entrepreneurs are different from the risk and they usually believe in something that others cannot see.
The interesting point is #1, "Transformational Value". Nick defines value as Benefit divided by Costs. Everybody knows that. But he adds something else to the equation, factor all that by the Benefits/Cost of your competitor to see the Transformational value.
If competitor A has a product that is worth $5 and lasts 10 years, the value of that is 2 (10 / $5). If your product costs $10 and lasts 20 years with everything else being the same, your value is also 2 (20 / $10). If you divide your value by your competitors (2 / 2) you get 1, which means you didn't create any transformational value.
He thinks that Startups should be looking into 10 to 100 times more value to be worth the trouble. Yeah, I agree to a certain extent.
He is missing something...
Nick clearly defines Transformational Value, but IMHO he is missing an important aspect of when to decide to invest in a company (or start a company, which is a type of investment). The missing parameter is Risk.
If the Transformational Value (TV) is 1 or less, that is not worth the trouble at all. If the TV is between 2 and 10, it might be a worthwhile investment if the risk is near zero. That is, if you can't loose and you will create some additional value, why not?
VCs like him are more interested in high-risk high-reward deals, and that is fine. However a lot of good people watched that presentation and saw that they couldn't create (or justify) a 10x Transformation Value. Should they quit? Shouldn't they try? Maybe they will fail, and that will be a great learning opportunity, or, maybe, they will find a niche or a morphed product that will hit big.
Jonathan Roberts (a.k.a. Jon Roberts) from Ignition Partners gave a talk on how to scale your business. The first thing that he said is to not scale too early, and then he listed 9 things that should be done/achieved before starting scaling, which are:
Determine your main value-add;
Evaluate the market and determine who your customer is and what they want to buy;
Determine your competitive strategy;
Determine your positioning;
Identify the key trends you're going to exploit;
Produce the product or a proto-type of the product at minimal cost and sell it to a few customers;
Fully deploy with a handful of customers;
Acquire the next handful of customers primarily through word of mouth;
Use PR and momentum to project your product, offering, or company into the market;
Scale.
Pretty fantastic list, but the terminology applies a bit more to an Enterprise-like product, not so much for a consumer product. For example, "Fully deploy with a handful of customers" in the consumer space is more like "Fully deploy/get used by a few thousand customers".
At Sampa...
Looking at Sampa and everything that we've done or achieved so far, I say we are pretty good at understanding and developing the business. Here is a point-by-point:
Value-add: We understand our value add.
We have a broad definition of customer that we are narrowing, narrowing, narrowing.
Competitive Strategy: It is a crowded space, but we have a decent strategy here.
Positioning: Ouch. Don't touch that, it hurts. Yes, we had a "Universe" positioning, then we had a "Dot" positioning, now we have (or are working on) an "Area" positioning.
Ahhh... We know that very well. We even got a lot of press because of that.
Produce the product: Done.
Deploy to customers: Done.
WOM: We do a little bit, but we want to do it better. Viraly.
Subscribe to this Site
. Guy summarizes very good points from the book.
October, 2007